NSF commits $76 million to near-market wireless projects, more than double its target

The National Science Foundation (NSF) planned to fund up to 10 teams to turn next-generation wireless research into deployable technology. It ended up funding 17, committing about $76 million against a $33 million target, according to an R&D World analysis of NSF’s award database. Its basic research track, by contrast, came in under its funding target, even with two more projects than the 30 it planned.

The awards land as NSF’s overall grantmaking shrinks. The agency was on track to issue the fewest new grants in four decades in fiscal 2026. The White House has requested a 55% cut for fiscal 2027 and House appropriators have proposed 20%, though Congress has rejected similar proposed cuts in the past.

NSF announced Oct. 16 that it has awarded more than $110 million to 49 teams under Verticals-enabling Intelligent NEtwork Systems (VINES), its new next-generation (NextG) wireless program, roughly three times what its predecessor, Resilient and Intelligent Next-Generation Systems (RINGS), awarded in 2022.

NSF first announced the program on June 3, 2025, saying it would invest up to $100 million “to accelerate performance and capabilities of next-generation (NextG) advanced intelligent network systems spanning the user-edge-core-cloud continuum.” The program is split into two tracks, one for use-inspired fundamental research (Track 1) and one for higher-readiness technology development and demonstration (Track 2).

In its solicitation, NSF said it was targeting 20 to 30 Track 1 awards up to $1.5 million each and 8 to 10 Track 2 awards up to $6 million each. Ultimately, NSF said it funded 32 Track 1 projects and 17 Track 2 projects. The award records total about $28.5 million for Track 1 and $75.9 million for Track 2, or roughly $104 million. An NSF spokesperson said in an email to R&D World that the more than $110 million figure combines NSF funding with $4.5 million from Ericsson, Intel and Qualcomm ($1.5 million each) for Track 1 and $1.95 million from federal partners for Track 2. It does not include international contributions, which fund international teams directly. No NSF money goes to those teams, a spokesperson said.

NSF received “significantly more Track 2 proposals than were anticipated and many of the proposals were deemed of high quality and worthy of funding,” a spokesperson said. The counts in the solicitation “were intended only as estimates, subject to availability of funding and the quality of proposals received.”

Ericsson is also a Track 2 awardee, slated to receive up to about $637,000 as part of a team developing wireless networks for extended reality. The solicitation says industry contributions will not fund Track 2 projects. NSF runs the two tracks separately, a spokesperson said, with its Computer and Information Science and Engineering (CISE) directorate managing Track 1 and its Technology, Innovation and Partnerships (TIP) directorate managing Track 2, each with its own proposal intake, review, decisions and funding. Track 1 industry partners cannot apply for Track 1 funding, and when they apply for Track 2, they “have no visibility into Track 2 competition or program management,” a spokesperson said.

NSF has obligated about $43 million so far, according to the award database. A spokesperson confirmed that funding for later years of the awards depends on future appropriations.

Disaster response and aerial networks lead Track 2

At least 14 Track 1 projects reference AI, machine learning or large language models in their titles. Other focuses include quantum, hardware, security and space. Most awardees are universities, though Track 2 has eight companies, including Ericsson and Nokia.

Under Track 2, four teams working on disaster and emergency response are slated to receive up to $18.5 million combined. Three other teams are working on aerial and aviation networks. Other focus areas include agriculture, road safety, AI-native networking and radio access networks for satellites.

A Northeastern University team will receive up to $5.1 million over three years, the largest single-institution award, for its “Untethering AI Factories” project. The team aims to replace some of the cabling in AI data centers with high-speed sub-terahertz wireless links, focusing on the separate network operators use to monitor and control servers. According to the award abstract, as AI clusters scale beyond 100,000 processors, that cabling becomes “a critical bottleneck.” It plans to test the system at an operating AI data center. Teradio Inc., a company developing wireless infrastructure for AI facilities, will lead commercialization.